Transplant International ISSN 0934-0874
ORIGINAL ARTICLE
Neurological complications after liver transplantation as a consequence of immunosuppression: univariate and multivariate analysis of risk factors Gianluca Rompianesi,1 Roberto Montalti,1 Nicola Cautero,1 Nicola De Ruvo,1 Anthony Stafford,2 Carolina Bronzoni,3 Roberto Ballarin,1 Lesley De Pietri,4 Fabrizio Di Benedetto1 and Giorgio E. Gerunda3 1 2 3 4
Liver Transplant Centre, Azienda Ospedaliero-Universitaria di Modena Policlinico, Modena, Italy Institute of Liver Studies, King’s College Hospital, London, UK General Surgery, Azienda Ospedaliero-Universitaria di Modena Policlinico, Modena, Italy Division of Anaesthesiology and Intensive Care Unit, Azienda Ospedaliero-Universitaria di Modena Policlinico, Modena, Italy
Keywords calcineurin inhibitors, immunosuppression, liver transplantation, mTOR inhibitors, neurological complications. Correspondence Gianluca Rompianesi MD, Liver Transplant Centre, Azienda Ospedaliero-Universitaria di Modena Policlinico, via del Pozzo, 71, 41124 Modena, Italy. Tel.: +39 059 4225265; fax: +39 059 4223765; e-mail: [email protected] Conflicts of interest The authors did not receive any conflict of interests that may bias the present study. Received: 18 November 2014 Revision requested: 30 December 2014 Accepted: 15 March 2015 Published online: 6 April 2015
Summary Neurological complications (NCs) can frequently and significantly affect morbidity and mortality of liver transplant (LT) recipients. We analysed incidence, risk factors, outcome and impact of the immunosuppressive therapy on NC development after LT. We analysed 478 LT in 440 patients, and 93 (19.5%) were followed by NCs. The average LOS was longer in patients experiencing NCs. The 1-, 3- and 5-year graft survival and patient survival were similar in patients with or without a NC. Multivariate analysis showed the following as independent risk factors for NC: a MELD score ≥20 (OR = 1.934, CI = 1.186–3.153) and an immunosuppressive regimen based on calcineurin inhibitors (CNIs) (OR = 1.669, CI = 1.009–2.760). Among patients receiving an everolimus-based immunosuppression, the 7.1% developed NCs, vs. the 16.9% in those receiving a CNI (P = 0.039). There was a 1-, 3- and 5-year NC-free survival of 81.7%, 81.1% and 77.7% in patients receiving a CNI-based regimen and 95.1%, 93.6% and 92.7% in those not receiving a CNI-based regimen (P < 0.001). In patients undergoing a LT and presenting with nonmodifiable risk factors for developing NCs, an immunosuppressive regimen based on CNIs is likely to result in a higher rate of NCs compared to mTOR inhibitors.
doi:10.1111/tri.12564
Introduction Despite the constant advances in organ preservation, immunosuppression, surgical techniques and perioperative management, several complications may occur after liver transplantation (LT) and affect the recipient’s postsurgical outcome. Neurologic complications (NC) can frequently and significantly affect morbidity and mortality and prolong the hospital length of stay (LOS) of LT recipients, therefore influencing the overall management costs. Such 864
complications usually occur in the first weeks after LT [1], showing prevalence in the literature of 11 to >40% [1–4]. Several pre-LT, surgical and postoperative factors may contribute to the development of NCs. It is postulated that among these factors, the side effects of the immunosuppressive drugs may play a leading role [5–8]. The aim of this study was to retrospectively analyse prevalence, risk factors, outcome and the impact of the immunosuppressive therapy on NC development following LT in 440 recipients. © 2015 Steunstichting ESOT 28 (2015) 864–869
Rompianesi et al.
Materials and methods From November 2000 to August 2012, we performed 478 LTs in 440 adult patients. The transplant was followed by a NC in 93 cases (19.5%). NCs were diagnosed by a consultant neurologist using clinical examination and if necessary additional investigations such as CT, MRI, EEG, EMG. All patients were included in the statistical analysis irrespective of length of follow-up. The 93 transplants whose recipient suffered from a NC made up the NC group, leaving the remaining 385 transplants as the control group. The number of transplants rather than patients was used for the analysis because 43 patients (9%) required more than one transplant. Requiring more than one transplant was included as a parameter in the analysis of risk factors. We considered in the analysis both major and minor NCs. Minor complications are usually self-limiting and do not require intervention or only a symptomatic treatment (tremors, mood alterations, sensory peripheral neuropathy, restless leg syndrome); all other NCs were considered as ‘major’: seizures, central pontine myelinolysis, flaccid paralysis, consciousness alterations, and toxic and metabolic encephalopathies [9,10]. Sensory peripheral neuropathy was defined as the presence of paraesthesia, dysaesthesia, allodynia or burning sensation in hands or feet, and an EMG was performed when indicated by the consultant neurologist. No patients were lost to follow-up, which generally consisted on a monthly ambulatory examination during the first year, four times during the second year, twice during the third year and annually thereafter. Factors analysed in the univariate analysis included patient’s pre-LT characteristics such as patient age, gender, first/redo LT, indication, isolated/combined liver–kidney transplantation, presence of encephalopathy and model of end-stage liver disease (MELD) score; surgical variables such as graft ischaemia time, length of surgery and intraoperative hypotensive episodes; donor variables such as age, gender and percentage of steatosis; and finally the immunosuppressive regimen. We also analysed the overall hospital LOS and the intensive care unit LOS, the patient outcome after the NC and the graft and patient survival. The immunosuppressive therapy used during the first months after LT was based on calcineurin inhibitors (CNI) such as cyclosporine or tacrolimus, or mammalian target of rapamycin inhibitors (mTOR) such as rapamycin or everolimus. Trough level analysis compared the blood trough levels of the immunosuppressant drugs at the time of the NC diagnosis for patients in the NC group, with the average trough level for patients in the control group. Limitations of this study are the retrospective analysis of data and that the patients were not randomized to receive the different immunosuppressive drugs. © 2015 Steunstichting ESOT 28 (2015) 864–869
Neurological complications after LT
In the analysis of risk factors, we considered the immunosuppressive regimen that the patients received before the development of the NC or in the first 3 months after the LT. The possible treatments given to patients with NCs were classified as follows: no treatment, immunosuppressive switch from CNI to mTOR inhibitors, adjustment of the daily dose of the immunosuppressive drugs to the lower limit of the therapeutic range or ‘symptomatic treatment’. This was when additional medications were prescribed to treat the NCs such as antiepileptic drugs, antipsychotic drugs, benzodiazepine, gamma-aminobutyric acid (GABA) analogues and B-complex vitamins. Statistical analysis Continuous data were reported as mean standard deviation for the parameters with a normal distribution or median (range) for parameters without a normal distribution and were compared using the two-sided Student’s t-test. Comparisons between groups for categorical variables were carried out using the chi-square test, with Yates’ correction, or Fisher’s exact test when appropriate. The optimal cut-off of continuous variables was obtained using the receiver operating characteristic (ROC) curve. Patient survival was evaluated using the Kaplan–Meier method and compared with the log-rank test. All variables with P < 0.05 in the univariate analysis were included in a multivariate logistic regression analysis by forward stepwise and backward stepwise method. The P-value of 0.05 or less was considered statistically significant. All analyses were carried out using SPSS 19.0 for Windows (SPSS, Chicago, IL, USA). Results Mean follow-up for all patients was 53.4 34.8 months, 48.9 34.7 months in the NC group and 55.1 34.8 in the control group (P = 0.124). The NCs occurred in a median time of 0.3 months (range 0–59.7 months) after LT. Among the 93 NCs observed, 49 of them (52.7%) were major, whilst 44 (47.3%) were minor (Table 1). In 30 patients (32.3%), the NCs have not been treated; in 23 patients (24.7%), an immunosuppressive switch from CNIs to mTOR inhibitors has been performed; in 11 patients (11.8%), the treatment consisted on an adjustment of the daily dose of the immunosuppressive drugs to the lower cut-off of the therapeutic range; and in 29 patients (31.2%), symptomatic treatment has been prescribed after assessment by a neurologist. Specifically, of the 50 major NCs, 27 cases (54%) had their immunosuppressive regimen modified and the remaining cases were treated symptomatically. Regarding the outcome of patients, 78 of the 93 patients who experienced a NC (83.9%) recovered fully from the NC, two patients died (2.2%) 4 and 11 days after 865
Rompianesi et al.
Neurological complications after LT
Table 1. Classification of the neurological complications. Neurological complications (n = 93)
No. (%)
Major complications Seizures* Consciousness alterations Toxic/metabolic encephalopathy Central pontine myelinolysis Flaccid paralysis Minor complications Sensory peripheral neuropathy Tremors Mood alterations Restless leg syndrome
49 (52.7) 17 (18.3) 13 (14) 12 (12.9) 6 (6.5) 1 (1.1) 44 (47.3) 30 (32.3) 9 (9.7) 4 (4.3) 1 (1.1)
*In one patient, the seizures were a consequence of disseminated aspergillosis with brain abscesses.
the diagnosis of the NC due to multiple organ failure and sepsis, respectively. The average hospital LOS of those who experienced a NC was significantly longer (26.2 16.1 days vs. 18.1 16.1 days; P < 0.001), especially in patients developing a major NC: 31.5 28.5 days vs. 20.6 10.1 days in patients developing a minor NC. The 1-, 3- and 5-year patient survival of patients who developed a NC was 78.2%, 70.5% and 70.5% vs. 83.4%, 75.0% and 69.1% (P = 0.662, Fig. 1). The 1-, 3- and 5-year graft survival of patients who developed a NC was 77.1%, 70.6% and 70.6% vs. 78.2%, 69.4% and 64.1% (P = 0.443, Fig. 1). The 1-, 3- and 5-year patient survival of patients who developed a major NC was 72.8%, 67.2% and 67.2% vs. 84.0%, 74.1% and 74.1% in patients who developed a minor NC (P = 0.626). The 1-, 3- and 5-year graft survival of patients who developed a major NC was 72.9%, 70.1% and 67.0% vs. 81.7%, 74.2% and 74.2% in patients who developed a minor NC (P = 0.724). Among the risk factors analysed in 93 transplants, a MELD score ≥20 (P = 0.012), the presence of pre-LT
encephalopathy (P = 0.009) and a CNI-based immunosuppression therapy (P = 0.024) were independent variables that help identify those at risk of developing NCs. Analysing separately the variables of the MELD score (serum total bilirubin, serum creatinine and INR), only a serum bilirubin level ≥4 mg/dl (68.4 lmol/l) was related to an higher incidence of NC (P < 0.001). All the factors analysed are summarized in Table 2. Multivariate logistic regression analysis showed that the variables independently associated with an increased risk of developing a NC were a MELD score ≥20 (OR = 1.934, CI = 1.186–3.153) and a CNI-based immunosuppression regimen (OR = 1.669, CI = 1.009–2.760). Of those who received everolimus immunosuppression, 7.1% of them developed a NC, compared with 21.7% of patients who received a CNI-based immunosuppression (P = 0.06). All the NCs of the five patients receiving everolimus were minor: two sensory peripheral neuropathy, two tremors and one mood alterations. There was no significant difference in the average immunosuppressive drug blood trough levels between the two groups. Specifically, the average blood trough levels were as follows: 9.2 3.4 vs. 9.3 2.6 ng/ml (P = 0.936) for everolimus, 8.7 3.6 vs. 8.4 2.9 ng/ml (P = 0.856) for rapamycin, 9.2 3.7 vs. 8.4 2.8 ng/ml (P = 0.207) for tacrolimus and 234 42 vs. 225 34 ng/ml (P = 0.109) for CsA in patients in the NC group and control group, respectively. The 1-, 3- and 5-year NC-free survival was 81.7%, 81.1% and 77.7% in patients receiving a CNI-based immunosuppression and 95.1%, 93.6% and 92.7% in patients receiving a CNI-free immunosuppression (P < 0.001, Fig. 2). Discussion Patient survival after LT dramatically improved since 1963 when Thomas Starzl performed the first LT in humans
Figure 1 Overall patient and graft survival after LT in patients with and without neurological complications (NC).
866
© 2015 Steunstichting ESOT 28 (2015) 864–869
Rompianesi et al.
Neurological complications after LT
Table 2. Univariate analysis of risk factors for developing a neurological complication (NC). Statistically significant values ( p < 0.05) highlighted in bold.
Recipient age Gender Pre-LT encephalopathy Primary transplant Combined transplant Living donor LT UNOS status 1 Indication
HCC Cold ischaemia time Length of surgery Hypotensive episodes Donor age MELD score
MELD score ≥20 Serum bilirubin ≥4 mg/dl Administration of CNI
Years M/F Yes/no Yes/no Yes/no Yes/no Yes/no Alcohol HBV/HCV Cholestatic Yes/no Minutes Minutes Yes/no Years Value Bilirubin (mg/dl) Creatinine (mg/dl) INR (value) Yes/no Yes/no Yes/no
NC group (93 transplants)
Control group (385 transplants)
P-value
51.8 11.1 65 (18.5%)/28 (22.0%) 53 (25.0%)/40 (15.0%) 80 (18.8%)/13 (24.5%) 6 (25.0%)/87 (19.2%) 8 (21.6%)/85 (19.3%) 10 (20.0%)/83 (19.4%) 12 (23.1%)/81 (19.0%) 61 (19.5%)/32 (18.7%) 15 (13.0%)/78 (17.2%) 35 (17.6%)/58 (20.8%) 344 117 509 170 5 (20.1%)/88 (19.4%) 53 19 22.3 9.4 11.2 11.5 1.4 1.1 2.0 7.2 44 (25.9%)/49 (15.9%) 42 (25%)/51 (16.5%) 82 (21.7%)/11 (11.0%)
52.2 11.1 286 (81.5%)/99 (78.0%) 159 (75.0%)/226 (85.0%) 345 (81.2%)/40 (75.5%) 18 (75.0%)/367 (80.8%) 29 (79.4%)/356 (80.7%) 40 (80.0%)/345 (80.6%) 40 (77.9%)/345 (81.0%) 246 (80.5%)/139 (81.3%) 110 (87.0%)/375 (82.8%) 164 (82.4%)/221 (79.2%) 345 118 492 120 19 (79.9%)/366 (80.6%) 56 19 18.6 9.2 6.6 11.5 1.2 1.1 2.1 7.2 126 (74.1%)/259 (84.1%) 126 (75%)/259 (83.5%) 296 (78.3%)/89 (89.0%)
0.755 0.465 0.009 0.421 0.660 0.896 0.931 0.608 0.853 0.205 0.451 0.941 0.263 0.929 0.0172
ORIGINAL ARTICLE
Neurological complications after liver transplantation as a consequence of immunosuppression: univariate and multivariate analysis of risk factors Gianluca Rompianesi,1 Roberto Montalti,1 Nicola Cautero,1 Nicola De Ruvo,1 Anthony Stafford,2 Carolina Bronzoni,3 Roberto Ballarin,1 Lesley De Pietri,4 Fabrizio Di Benedetto1 and Giorgio E. Gerunda3 1 2 3 4
Liver Transplant Centre, Azienda Ospedaliero-Universitaria di Modena Policlinico, Modena, Italy Institute of Liver Studies, King’s College Hospital, London, UK General Surgery, Azienda Ospedaliero-Universitaria di Modena Policlinico, Modena, Italy Division of Anaesthesiology and Intensive Care Unit, Azienda Ospedaliero-Universitaria di Modena Policlinico, Modena, Italy
Keywords calcineurin inhibitors, immunosuppression, liver transplantation, mTOR inhibitors, neurological complications. Correspondence Gianluca Rompianesi MD, Liver Transplant Centre, Azienda Ospedaliero-Universitaria di Modena Policlinico, via del Pozzo, 71, 41124 Modena, Italy. Tel.: +39 059 4225265; fax: +39 059 4223765; e-mail: [email protected] Conflicts of interest The authors did not receive any conflict of interests that may bias the present study. Received: 18 November 2014 Revision requested: 30 December 2014 Accepted: 15 March 2015 Published online: 6 April 2015
Summary Neurological complications (NCs) can frequently and significantly affect morbidity and mortality of liver transplant (LT) recipients. We analysed incidence, risk factors, outcome and impact of the immunosuppressive therapy on NC development after LT. We analysed 478 LT in 440 patients, and 93 (19.5%) were followed by NCs. The average LOS was longer in patients experiencing NCs. The 1-, 3- and 5-year graft survival and patient survival were similar in patients with or without a NC. Multivariate analysis showed the following as independent risk factors for NC: a MELD score ≥20 (OR = 1.934, CI = 1.186–3.153) and an immunosuppressive regimen based on calcineurin inhibitors (CNIs) (OR = 1.669, CI = 1.009–2.760). Among patients receiving an everolimus-based immunosuppression, the 7.1% developed NCs, vs. the 16.9% in those receiving a CNI (P = 0.039). There was a 1-, 3- and 5-year NC-free survival of 81.7%, 81.1% and 77.7% in patients receiving a CNI-based regimen and 95.1%, 93.6% and 92.7% in those not receiving a CNI-based regimen (P < 0.001). In patients undergoing a LT and presenting with nonmodifiable risk factors for developing NCs, an immunosuppressive regimen based on CNIs is likely to result in a higher rate of NCs compared to mTOR inhibitors.
doi:10.1111/tri.12564
Introduction Despite the constant advances in organ preservation, immunosuppression, surgical techniques and perioperative management, several complications may occur after liver transplantation (LT) and affect the recipient’s postsurgical outcome. Neurologic complications (NC) can frequently and significantly affect morbidity and mortality and prolong the hospital length of stay (LOS) of LT recipients, therefore influencing the overall management costs. Such 864
complications usually occur in the first weeks after LT [1], showing prevalence in the literature of 11 to >40% [1–4]. Several pre-LT, surgical and postoperative factors may contribute to the development of NCs. It is postulated that among these factors, the side effects of the immunosuppressive drugs may play a leading role [5–8]. The aim of this study was to retrospectively analyse prevalence, risk factors, outcome and the impact of the immunosuppressive therapy on NC development following LT in 440 recipients. © 2015 Steunstichting ESOT 28 (2015) 864–869
Rompianesi et al.
Materials and methods From November 2000 to August 2012, we performed 478 LTs in 440 adult patients. The transplant was followed by a NC in 93 cases (19.5%). NCs were diagnosed by a consultant neurologist using clinical examination and if necessary additional investigations such as CT, MRI, EEG, EMG. All patients were included in the statistical analysis irrespective of length of follow-up. The 93 transplants whose recipient suffered from a NC made up the NC group, leaving the remaining 385 transplants as the control group. The number of transplants rather than patients was used for the analysis because 43 patients (9%) required more than one transplant. Requiring more than one transplant was included as a parameter in the analysis of risk factors. We considered in the analysis both major and minor NCs. Minor complications are usually self-limiting and do not require intervention or only a symptomatic treatment (tremors, mood alterations, sensory peripheral neuropathy, restless leg syndrome); all other NCs were considered as ‘major’: seizures, central pontine myelinolysis, flaccid paralysis, consciousness alterations, and toxic and metabolic encephalopathies [9,10]. Sensory peripheral neuropathy was defined as the presence of paraesthesia, dysaesthesia, allodynia or burning sensation in hands or feet, and an EMG was performed when indicated by the consultant neurologist. No patients were lost to follow-up, which generally consisted on a monthly ambulatory examination during the first year, four times during the second year, twice during the third year and annually thereafter. Factors analysed in the univariate analysis included patient’s pre-LT characteristics such as patient age, gender, first/redo LT, indication, isolated/combined liver–kidney transplantation, presence of encephalopathy and model of end-stage liver disease (MELD) score; surgical variables such as graft ischaemia time, length of surgery and intraoperative hypotensive episodes; donor variables such as age, gender and percentage of steatosis; and finally the immunosuppressive regimen. We also analysed the overall hospital LOS and the intensive care unit LOS, the patient outcome after the NC and the graft and patient survival. The immunosuppressive therapy used during the first months after LT was based on calcineurin inhibitors (CNI) such as cyclosporine or tacrolimus, or mammalian target of rapamycin inhibitors (mTOR) such as rapamycin or everolimus. Trough level analysis compared the blood trough levels of the immunosuppressant drugs at the time of the NC diagnosis for patients in the NC group, with the average trough level for patients in the control group. Limitations of this study are the retrospective analysis of data and that the patients were not randomized to receive the different immunosuppressive drugs. © 2015 Steunstichting ESOT 28 (2015) 864–869
Neurological complications after LT
In the analysis of risk factors, we considered the immunosuppressive regimen that the patients received before the development of the NC or in the first 3 months after the LT. The possible treatments given to patients with NCs were classified as follows: no treatment, immunosuppressive switch from CNI to mTOR inhibitors, adjustment of the daily dose of the immunosuppressive drugs to the lower limit of the therapeutic range or ‘symptomatic treatment’. This was when additional medications were prescribed to treat the NCs such as antiepileptic drugs, antipsychotic drugs, benzodiazepine, gamma-aminobutyric acid (GABA) analogues and B-complex vitamins. Statistical analysis Continuous data were reported as mean standard deviation for the parameters with a normal distribution or median (range) for parameters without a normal distribution and were compared using the two-sided Student’s t-test. Comparisons between groups for categorical variables were carried out using the chi-square test, with Yates’ correction, or Fisher’s exact test when appropriate. The optimal cut-off of continuous variables was obtained using the receiver operating characteristic (ROC) curve. Patient survival was evaluated using the Kaplan–Meier method and compared with the log-rank test. All variables with P < 0.05 in the univariate analysis were included in a multivariate logistic regression analysis by forward stepwise and backward stepwise method. The P-value of 0.05 or less was considered statistically significant. All analyses were carried out using SPSS 19.0 for Windows (SPSS, Chicago, IL, USA). Results Mean follow-up for all patients was 53.4 34.8 months, 48.9 34.7 months in the NC group and 55.1 34.8 in the control group (P = 0.124). The NCs occurred in a median time of 0.3 months (range 0–59.7 months) after LT. Among the 93 NCs observed, 49 of them (52.7%) were major, whilst 44 (47.3%) were minor (Table 1). In 30 patients (32.3%), the NCs have not been treated; in 23 patients (24.7%), an immunosuppressive switch from CNIs to mTOR inhibitors has been performed; in 11 patients (11.8%), the treatment consisted on an adjustment of the daily dose of the immunosuppressive drugs to the lower cut-off of the therapeutic range; and in 29 patients (31.2%), symptomatic treatment has been prescribed after assessment by a neurologist. Specifically, of the 50 major NCs, 27 cases (54%) had their immunosuppressive regimen modified and the remaining cases were treated symptomatically. Regarding the outcome of patients, 78 of the 93 patients who experienced a NC (83.9%) recovered fully from the NC, two patients died (2.2%) 4 and 11 days after 865
Rompianesi et al.
Neurological complications after LT
Table 1. Classification of the neurological complications. Neurological complications (n = 93)
No. (%)
Major complications Seizures* Consciousness alterations Toxic/metabolic encephalopathy Central pontine myelinolysis Flaccid paralysis Minor complications Sensory peripheral neuropathy Tremors Mood alterations Restless leg syndrome
49 (52.7) 17 (18.3) 13 (14) 12 (12.9) 6 (6.5) 1 (1.1) 44 (47.3) 30 (32.3) 9 (9.7) 4 (4.3) 1 (1.1)
*In one patient, the seizures were a consequence of disseminated aspergillosis with brain abscesses.
the diagnosis of the NC due to multiple organ failure and sepsis, respectively. The average hospital LOS of those who experienced a NC was significantly longer (26.2 16.1 days vs. 18.1 16.1 days; P < 0.001), especially in patients developing a major NC: 31.5 28.5 days vs. 20.6 10.1 days in patients developing a minor NC. The 1-, 3- and 5-year patient survival of patients who developed a NC was 78.2%, 70.5% and 70.5% vs. 83.4%, 75.0% and 69.1% (P = 0.662, Fig. 1). The 1-, 3- and 5-year graft survival of patients who developed a NC was 77.1%, 70.6% and 70.6% vs. 78.2%, 69.4% and 64.1% (P = 0.443, Fig. 1). The 1-, 3- and 5-year patient survival of patients who developed a major NC was 72.8%, 67.2% and 67.2% vs. 84.0%, 74.1% and 74.1% in patients who developed a minor NC (P = 0.626). The 1-, 3- and 5-year graft survival of patients who developed a major NC was 72.9%, 70.1% and 67.0% vs. 81.7%, 74.2% and 74.2% in patients who developed a minor NC (P = 0.724). Among the risk factors analysed in 93 transplants, a MELD score ≥20 (P = 0.012), the presence of pre-LT
encephalopathy (P = 0.009) and a CNI-based immunosuppression therapy (P = 0.024) were independent variables that help identify those at risk of developing NCs. Analysing separately the variables of the MELD score (serum total bilirubin, serum creatinine and INR), only a serum bilirubin level ≥4 mg/dl (68.4 lmol/l) was related to an higher incidence of NC (P < 0.001). All the factors analysed are summarized in Table 2. Multivariate logistic regression analysis showed that the variables independently associated with an increased risk of developing a NC were a MELD score ≥20 (OR = 1.934, CI = 1.186–3.153) and a CNI-based immunosuppression regimen (OR = 1.669, CI = 1.009–2.760). Of those who received everolimus immunosuppression, 7.1% of them developed a NC, compared with 21.7% of patients who received a CNI-based immunosuppression (P = 0.06). All the NCs of the five patients receiving everolimus were minor: two sensory peripheral neuropathy, two tremors and one mood alterations. There was no significant difference in the average immunosuppressive drug blood trough levels between the two groups. Specifically, the average blood trough levels were as follows: 9.2 3.4 vs. 9.3 2.6 ng/ml (P = 0.936) for everolimus, 8.7 3.6 vs. 8.4 2.9 ng/ml (P = 0.856) for rapamycin, 9.2 3.7 vs. 8.4 2.8 ng/ml (P = 0.207) for tacrolimus and 234 42 vs. 225 34 ng/ml (P = 0.109) for CsA in patients in the NC group and control group, respectively. The 1-, 3- and 5-year NC-free survival was 81.7%, 81.1% and 77.7% in patients receiving a CNI-based immunosuppression and 95.1%, 93.6% and 92.7% in patients receiving a CNI-free immunosuppression (P < 0.001, Fig. 2). Discussion Patient survival after LT dramatically improved since 1963 when Thomas Starzl performed the first LT in humans
Figure 1 Overall patient and graft survival after LT in patients with and without neurological complications (NC).
866
© 2015 Steunstichting ESOT 28 (2015) 864–869
Rompianesi et al.
Neurological complications after LT
Table 2. Univariate analysis of risk factors for developing a neurological complication (NC). Statistically significant values ( p < 0.05) highlighted in bold.
Recipient age Gender Pre-LT encephalopathy Primary transplant Combined transplant Living donor LT UNOS status 1 Indication
HCC Cold ischaemia time Length of surgery Hypotensive episodes Donor age MELD score
MELD score ≥20 Serum bilirubin ≥4 mg/dl Administration of CNI
Years M/F Yes/no Yes/no Yes/no Yes/no Yes/no Alcohol HBV/HCV Cholestatic Yes/no Minutes Minutes Yes/no Years Value Bilirubin (mg/dl) Creatinine (mg/dl) INR (value) Yes/no Yes/no Yes/no
NC group (93 transplants)
Control group (385 transplants)
P-value
51.8 11.1 65 (18.5%)/28 (22.0%) 53 (25.0%)/40 (15.0%) 80 (18.8%)/13 (24.5%) 6 (25.0%)/87 (19.2%) 8 (21.6%)/85 (19.3%) 10 (20.0%)/83 (19.4%) 12 (23.1%)/81 (19.0%) 61 (19.5%)/32 (18.7%) 15 (13.0%)/78 (17.2%) 35 (17.6%)/58 (20.8%) 344 117 509 170 5 (20.1%)/88 (19.4%) 53 19 22.3 9.4 11.2 11.5 1.4 1.1 2.0 7.2 44 (25.9%)/49 (15.9%) 42 (25%)/51 (16.5%) 82 (21.7%)/11 (11.0%)
52.2 11.1 286 (81.5%)/99 (78.0%) 159 (75.0%)/226 (85.0%) 345 (81.2%)/40 (75.5%) 18 (75.0%)/367 (80.8%) 29 (79.4%)/356 (80.7%) 40 (80.0%)/345 (80.6%) 40 (77.9%)/345 (81.0%) 246 (80.5%)/139 (81.3%) 110 (87.0%)/375 (82.8%) 164 (82.4%)/221 (79.2%) 345 118 492 120 19 (79.9%)/366 (80.6%) 56 19 18.6 9.2 6.6 11.5 1.2 1.1 2.1 7.2 126 (74.1%)/259 (84.1%) 126 (75%)/259 (83.5%) 296 (78.3%)/89 (89.0%)
0.755 0.465 0.009 0.421 0.660 0.896 0.931 0.608 0.853 0.205 0.451 0.941 0.263 0.929 0.0172
